Hyperparticle / udify

A single model that parses Universal Dependencies across 75 languages. Given a sentence, jointly predicts part-of-speech tags, morphology tags, lemmas, and dependency trees.
https://arxiv.org/abs/1904.02099
MIT License
219 stars 56 forks source link

"Training on other datasets" directions missing required flag #15

Closed neubig closed 4 years ago

neubig commented 4 years ago

This command in the "Training on other datasets" of the README causes the following error:

(python3) gneubig@ogma:~/work/udify$ python train.py --config config/ud/en/udify_bert_train_en_ewt.json --name en_ewt
Traceback (most recent call last):
  File "train.py", line 46, in <module>
    train_path = glob.glob(pathname).pop()
IndexError: pop from empty list

Adding the --dataset_dir resolves the error.

python train.py --config config/ud/en/udify_bert_train_en_ewt.json --name en_ewt --dataset_dir data/ud-treebanks-v2.3/
Hyperparticle commented 4 years ago

Good catch. Updated the README.